We can see uniform distribution of bikes rented across four seasons.
We see that maximum bikes are rented when the weather is clear and partly cloudy.
Uniform distribution can be seen at all times of the day.
Uniform distribution can be seen at all the days.
Uniform distribution can be seen across all the months.
Equal number of bikes were rented in both the years.
We can see a normal distribution of bike rented across the temperature.
We can see a normal distribution of bike rented across the atemp.
Here data is slightly skewed.
Data is skewed. Count of bikes rented decreases as windspeed increases. Count is maximum when windspeed is zero.
Count of non registered user rentals(entry) of zero is maximum at 986 days followed by count of one, two and three user rentals most of the times. 309 is the maximum number of registrations that has been done one day.
Count of registered user rentals(entry) of three is mazimum at 195 followed by four and five. 761 is the maximum number of registrations that has been done one day.
5 numer of bikes have been rented maximum number of days that is 169. The maximum number if bikes is 636 that has been rented one day.
There are few outliers in humidity data. This might be because of data entry errors or rare meteorological events can cause unusually high or low humidity levels.
There are few outliers in windspeed data. This might be because of data entry errors or extreme weather phenomena that can cause extremely high windspeed readings.
PREPROCESSING FOR THE MODEL We want to predict the best model to count number of bikes rented and find features that are most important and affect count of bikes rented.